评分规则承载标准:基于Shapley归因的清单式大语言模型(LLM)评判器
Rubrics Carry the Standard: Shapley Attribution for Checklist-Based LLM Judges
AI总结:
该研究针对清单式LLM评判器的两个假设,通过实验发现评分规则对通过率影响远大于评判器,提出Shapley归因方法控制自我审查导致的评分规则漂移,维持初始质量与一致决策。
摘要:
清单式评分规则被大语言模型(LLM)评判器采用,已成为评估开放式输出及训练时奖励输出的标准工具。围绕该实践存在两个假设:一是宽松或严格是评判器模型的属性,二是可通过让模型审查评分规则来改进它。我们在植入缺陷的受控评分任务及两个人工标注数据集SummEdits和RAGTruth上对这两个假设进行了测试。在固定评判器和输出的情况下,评分规则内容可使正确输出的通过率从2%变动至100%,而改变评判器模型或相同检查项的呈现方式仅能带来几个百分点的变动。模型自我审查会使评分规则偏向审查者自身的标准:它添加任务从未规定的要求,删除人工标注曾明确的检查项,在三个审查者模型及全部三个数据集上均如此。为测量和控制这一情况,我们证明在全通过聚合下,每个检查项的Shapley值存在精确闭式解,可在线性时间内得出每个检查项的捕获信用、误判责任及唯一捕获情况。归因分析显示,自我审查会删除具有唯一作用的检查项,保留导致误判的检查项。结合这些量与标注衍生检查项的来源冻结的变更控制门,可将几乎所有漂移的修订链维持在初始质量,在三个评判器模型下得出一致的决策。
英文摘要:
Checklist rubrics applied by large language model (LLM) judges have become a standard instrument for evaluating open-ended outputs and for rewarding them during training. Two assumptions surround this practice: that leniency or strictness is a property of the judge model, and that a rubric can be improved by asking a model to review it. We test both in a controlled grading task with planted flaws and on two human-labeled datasets, SummEdits and RAGTruth. Holding the judge and the outputs fixed, rubric content moves the pass rate of correct outputs from 2% to 100%, whereas changing the judge model or the rendering of the same checks moves it by a few points. Model self-review then drifts the rubric toward the reviser's own standard: it adds requirements the task never stated and deletes checks that human labels had taught, with three reviser models and on all three datasets. To measure and control this, we show that under all-must-pass aggregation the Shapley value of each check has an exact closed form, which yields per-check catch credit, false-fail blame and unique catches in linear time. Attribution shows that self-review deletes the checks that do unique work and keeps the checks that cause false fails. A change-control gate that combines these quantities with a provenance freeze on label-derived checks holds nearly every drifting revision chain at its starting quality, with identical decisions under three judge models.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。